INTRODUCTION
Although writing has always been considered a vital component in the English as a foreign language (EFL) curriculum, EFL writing ability has gained even more prominence in the twenty-first century (Murcia, et al., 2014). Writing represents a fundamental communication skill that leaves a considerable impact on the process of learning a second language (L2) (Cárdenas, 2018). Mastering writing skill could bring about communicative, educational, and professional advantages (Yusuf et al., 2019). Writing-related abilities may appear in different modes/genres of texts, such as descriptive, narrative, expository, and argumentative, each of which could have its own influence on L2 writing performance, assessment, and linguistic features of texts (Esfandiari & Jafari, 2021). Among these, descriptive writing is concerned with the challenging task of outlining the details of an object or a setting through visualization (Schacter, 2015).
Like all modes of writing, one of the most serious issues in descriptive writing is how it should be assessed (Farhady, 2006). Although traditional modes of assessment can be used for this purpose, descriptive writing has been rarely assessed based on alternative frameworks including self- and peer assessment. As Bartlett (2015) observed, self-assessment is a technique through which learners can assess their own works and recognize their strengths/weaknesses. Peer assessment is another approach in which learners are allowed to assess each other’s works, while offering appropriate comments when needed (Falchikov, 2012).
Considering the novelty of peer and self-assessment of writing as opposed to traditional methods, a line of research has focused on learners’ attitudes towards this mode of assessment (e.g., Abolfazli-Khonbi & Sadeghi, 2013; Crusan, 2011; Suzuki, 2009). Given the importance of learner voice and attitude (Bloemert, et al., 2020), the attitudinal component in the assessment research helps evaluate the effectiveness of assessment, especially as it provides a platform through which EFL learners’ attitudes are directly taken into account. However, one of the major problems of peer and self-assessment techniques is that they often use either qualitative or quantitative designs, without employing attitudinal scales. Meanwhile, there is a shortage of studies exploring the impact of gender on EFL learners’ attitudes towards self- and peer assessment of descriptive writing, particularly in highly vibrant EFL contexts, such as Iranian EFL academies.
LITERATURE REVIEW
To date, many studies have been administered to investigate the field of assessment generally and peer and self-assessment specifically to add information to the body of research knowledge. In the following subsections, a few of which are presented.
Assessment in Language Teaching
The notion of assessment generally refers to the application of diverse instruments, approaches, and strategies in order to gather and interpret data regarding learners’ abilities and capabilities, through traditional quantitative measurement, testing, and other evaluative approaches (Mihai, 2010). Harmer (2015) believes that assessment is an integral segment of the process of teaching; hence, teachers should specifically focus on it. As such, this activity may result in motivational, regulative, prognostic, and developmental functions (Kolář & Šikulová, 2009).
Traditional assessment methods have relied on various forms of testing, such as true or false items, multiple-choice tests, and essay-type questions (Farhady, 2006; Simonson et al., 2003). Assessment practices, however, are not limited to traditional approaches, and they can be complemented by alternative assessment practices that underscore learners’ everyday activities (Esfandiari & Myford, 2013). Alternative assessment is distinguished from traditional assessment in many ways; for instance, (i) it does not impose classroom activities on students, (ii) it reflects the curriculum taken shape in the classroom, (iii) it provides information about each individual student’s strengths/weaknesses, and (iv) it is sensitive across various cultures and avoid norms or biases, e.g., cultural or linguistic forms which were problematic in the conventional types of language testing (Richards & Renandya, 2002).
Alternative assessment may appear in different forms, such as portfolio assessment, protocol analysis, journal entries, dialogue journals, peer assessment, and self-assessment. Among these possibilities, peer assessment and self-assessment offer both autonomous and structured assessment procedures to EFL learners. Self-assessment emphasizes the importance of learners’ ability to be responsible for various aspects of their own learning including setting learning goals, monitoring their own performance, taking learning decisions, and building their motivation (McNamara & Deane, 1995). Through self-assessment, students are engaged in deciding the standards/criteria that would determine their levels of learning (Boud, 2007). In this process of formative assessment, learners consider the quality of their work, assess the degree of their progress, and also try to improve their status based on the judgments (Gregory et al., 2011).
Another mode of alternative assessment is peer assessment. This notion rests on the ideas of theorists in social constructivism (e.g., Vygotsky) and in active learning (e.g., Piaget) (Falchikov 2012). Peer assessment is a technique by which students offer feedback to their peers based on predetermined criteria (Falchikov, 2012). Through this educational arrangement, a student judges a partner's performance qualitatively and/or quantitatively, which shapes a process that triggers students to express their reactions (e.g., agreement or dispute) in a cooperative setting (Strijbos & Sluijsmans, 2010). Peer assessment may be employed in different ways; for instance, Bartlett (2015) suggests Two Stars and A Wish as an effective strategy, in which pupils are asked to highlight two positives about their peers’ works and then to identify one area for improvement.
Numerous studies have explored self- and peer assessment from different angles. Assessment researchers have investigated the impact of self-assessment on teacher education, speaking skill, and EFL students' writing (e.g., Chen, 2008; Mok et al., 2006). Others have drawn upon students' perceptions and experiences of (formative) peer assessment (e.g., Kaufman & Shunn, 2010; Vickerman, 2009). Meanwhile, another line of research has focused on the effectiveness of peer and self-assessment for developing learners' independence and awareness (Birjandi & Hadidi-Tamjid, 2012). Probing into essay writing, Birjandi and Siyyari (2010) also asserted that peer assessment could have a great influence on EFL learners’ writing abilities.
Many researchers have emphasized the integrated application of peer and self-assessment. Black et al. (2003) highlighted the importance of peer assessment as an integral complement of self-assessment. Maiz-Arévalo (2008) stated that students exposed to peer assessment would most likely be able to assess their own productions in the future. This observation confirms that peer assessment could enhance self-assessment. According to Ashraf and Mahdinezhad (2015), self- and peer assessment techniques can reinforce learning and motivate students to be a part of a community (e.g., student community) to evaluate their partners’ tasks. Another crucial function of self- and peer assessment is the additional feedback provided by peers, which can help learners promote their autonomy in mastering language skills. At last, Hariri Asl and Marandi (2017) asserted that providing collaborative opportunity for students to learn from each other, rather than following their teachers’ comments only, is one of the major concerns among instructors and educational researchers. They also argued that peer assessment can enhance social and interpersonal skills and create ties and connections between feedback, receivers and presenters.
Attitude in Self- and Peer Assessment
The notion of attitude has been frequently used in studies exploring language learning (including ESL and EFL). An attitude is a constant series of thoughts and notions regarding a defined idea or a condition, and it can affect an individual’s preferences and reactions (Oroujlou & Vahedi, 2011). In fact, the relative endurance of attitude makes it possible for it to be learned and unlearned (Oroujlou & Vahedi, 2011). Plausibly, because an attitude can be learned, it can also be taught. Research in L2 learning suggests that positive attitudes could increase students’ efficiency in language learning. In fact, attitude is an important factor in influencing language performance and in eliciting EFL teachers' and learners' views towards language learning and teaching.
However, the literature on language proficiency (particularly writing) includes assessment techniques that have exclusively relied on prefabricated qualitative or quantitative frameworks to pass judgment on learners' writing ability. Meanwhile, integrated versions of self- and peer assessment have rarely been used, despite their effective functions. The results of some investigations have demonstrated the effects of self- and peer assessment in language assessment process on solving EFL issues, such as the impact of attitude and motivation on L2 learning (Gardner & Lambert, 1972), students' perceptions of ESL writing (Ismail, 2010), students' views toward peer, self-, and teacher assessment in relation to course achievement (Abolfazli-Khonbi & Sadeghi, 2013), attitude and friendship bias towards peer assessment in an EFL context (Azarnoosh, 2013), and students' attitude about how peer and self-assessment helped improve their experience of learning (Siow, 2015). These observations were found to emphasize the importance of investigating self- and peer assessment in the case of descriptive writing.
Descriptive Writing
Scholars in language teaching/learning view writing as a highly complex issue (Brown, 2004; Farhady, 2006; Weigle, 2002). In measuring writing ability, various analytic traits can be included, such as form, content, grammar, vocabulary, and mechanics (Farhady, 2006). These elements together constitute different genres of writing including descriptive, narrative, expository, and argumentative. Among these, descriptive writing remains as one of the under-researched areas in language teaching. Descriptive writing is a genre of writing that deals with the expression of the detailed characteristics of individuals, settings, or objects in a way that appeals to readers’ visualization ability (Schacter, 2015). As such, it involves the expression of sensory perceptions (e.g., visual, kinetic, auditory) (Kane, 2000).
Descriptive writing is considered a foundational type of writing task, because, as mentioned by Meyers (2009), most genres of writing consist of some elements of description. If writers illustrate an object, a person, or even a place with well-adjusted graphic information, they will be able to leave a memorable trace in their audiences’ minds (Suriyanti & Yaacob, 2016). From the perspective of education, it is convenient for EFL instructors and guidance school learners (Suriyanti & Yaacob, 2016). At the same time, as emphasized by Suriyanti and Yaacob (2016), it involves details that appeal to the five senses, along with figurative language (e.g., simile, personification, and metaphor). Details related to different senses can produce some vivid and evocative image that gives ‘life’ to words (Axelrod & Cooper, 2011).
Review of the literature revealed that most studies have investigated the application of self-assessment or peer assessment separately or concurrently in teacher education and ESL studies related to speaking and writing skills. However, such studies did not elicit their participants’ attitudes towards descriptive writing (e.g., Azarnoosh, 2013; Birjandi & Hadidi-Tamjid, 2012; Birjandi & Siyyari, 2010; Chen, 2008; Kaufman & Shunn, 2010; Mok et al., 2006; Vickerman, 2009). Additionally, most researchers investigating self- and peer assessment of writing has not explored a specific genre, such as descriptive writing. Definitely, descriptive writing can be further investigated through innovative methods and approaches; that explains why we decided to explored EFL learners' attitudes towards self- and peer assessment of descriptive writing in an Iranian EFL context through exploratory sequential mixed-methods design.
PURPOSE OF THE STUDY
The overarching goal of this investigation was to elicit the attitudes of a specific number of Iranian EFL learners toward peer and self-assessment of descriptive writing. In doing so, the study relied on an exploratory sequential mixed-methods approach including qualitative and quantitative strands (Creswell & Plano Clark, 2017). Based on our knowledge, no investigation has ever been carried out to examine the current research topic; as such, this study attempted to fill the research niche through addressing the following research questions:
- What are the Iranian (female and male) EFL learners' views toward peer and self-assessment of descriptive writing?
- Do the factors identified predict the Iranian (female and male) EFL learners' views toward peer and self-assessment in the case of descriptive writing?
- Is there a statistically significant difference between the Iranian female and male EFL learners' views toward peer and self-assessment of descriptive writing based on the recognized themes?
METHOD
Research Design
The research design of the study included an exploratory sequential mixed-methods design (ESMMsD) developed by Creswell and Clark (2017) within the framework of Social Science Theory/SST. Hence, the summary of the design of the study is presented as follows (see Figure 1):
Social Science Theory /SST
|
Theories of Research Variables
|
Figure 1: Bottom-up representation of the theoretical framework of this explanatory sequential mixed-methods study
Figure 1 shows the schematic representation of the theoretical framework of this exploratory mixed-methods study; that is, throughout the research process, qualitative interview data is first gathered and then quantitative data by means of the questionnaire scores is collected. It can be claimed that ESMMsD assists us in the following ways. ESMMsD, first, helps the researchers in developing required interview and questionnaire for Iranian male and female EFL learners to have their views toward self- and peer-assessment of descriptive writing. Secondly, the sequential aspect of the design is in line with Bartlett's model of self- and peer assessment in making Iranian EFL learners autonomous, and in facilitating the burden of training of the learners to write a better descriptive composition both inside and outside of the classroom. Also, this design permits the researcher to recruit the same or different participants for the quantitative strand from the qualitative strand of the study. Following that, it contributes the researchers to conduct and collect only one strand of the study at a time and to increase the feasibility of the research, too. Lastly, this design helps the Iranian EFL teachers to teach self- and peer assessment of descriptive writing to their students sequentially, and provide comprehensive feedbacks for them.
Participants
To select the research participants, the purposive sampling procedure was used. The participants were as follows. In the first group, the participants comprised 20 male and 20 female EFL teachers who were teaching at foreign language institutes of Sanandaj (Iran). Among them, five volunteer teachers were selected as the finalized participants prior to the conduction of the main course of the study to train EFL elementary learners. The selected teachers were asked to take part in a training course related to self- and peer assessment of descriptive writing.
Furthermore, in the second group, the participants consisted of 50 EFL elementary learners who were equally divided into two groups (25 males and 25 females) from seven different language academies of Sanandaj, Iran, and aged between 11 and 14. It is worth noting that a descriptive topic was given to the learners and those persons whose compositions scores were below the mean value of the class (i.e., 10 out of 20) were selected purposefully because the researchers had no other choices to select another sampling procedure, due to the limited number of the participants in the selected language centers. Appendix A depicts the demographic data of the study in tabulated forms.
Data Collection Procedure
After determining research participants of this study, the researchers asked the teachers to participate in a training course, for at least 6 months in the Spring and Summer of 2019. In the first month of the Spring (April), all the selected Iranian EFL teachers were asked to prepare themselves to administer the course in two separate male and female only groups within 5 successive training sessions. As the course commenced, an overview of various sorts of assessment like peer and self-assessment and also several aspects of writing skills, such as narrative, expository, argumentative, and descriptive (composition) writing was given. With regard to self-assessment and peer assessment, short explanations, definitions, and their applications separately were introduced. With this, the participants became acquaintance with Bartlett's model (2015) of peer and self-assessment of descriptive (composition) writing. In the second and third sessions, the trainer (the first author) asked the teachers to write a composition (at least 100 words). Then, they were asked to deliver the papers to their classmates to assess them by means of the peer-assessment technique. Similar to peer assessment, they were asked to assess their own compositions through the self-assessment technique based on the information that they received from the training sessions.
Furthermore, the teachers practically became familiarized with sensory details, figurative language, and vivid words (SFV) model of Spencer (2005) in the fourth session. First, in the free writing stage, strategies of free-writing such as brainstorming and mind mapping were elucidated. Richards and Schmidt (2010) stated that brainstorming in writing instruction is a form of pre-writing that students can apply in expressing their ideas on a piece of paper without attending to word spelling, sentence structure, and writing mechanics. It also helps students to collect their ideas pertinent to an assigned topic and to have a better performance in writing. Richards and Schmidt (2010) went on to define mind mapping as a technique for categorizing new words or other learning content. A key word could provide a connection between related words and notions shown schematically. Secondly, the ‘whilst-writing’ stage of the SFV model was implemented and a detailed description of descriptive writing elements like figurative language, sensory details and vivid words were all provided. Suriyanti and Yaacob (2016) described sensory details as an amalgamation of the senses of human beings that permit individuals to see them as the real things. As Arp and Johnson (2006) stated, figurative language – language using figures of speech such as hyperbole, simile, metaphor, and personification – is a language that cannot be taken literally. Richards and Schmidt (2010) stated that hyperbole is a kind of exaggeration, e.g., her brain is the size of a pea. They defined simile as the application of function words such as like, than by which we compare and contrast one thing to another. They also stated that metaphor is a figure of speech when two items are parallel with one another, for instance, the man is a lion. According to Arp and Johnson (2006), personification helps to apply the traits of mankind to a thing, an animal, or a concept; for example, the sun smiled down on us. Researchers like Spencer (2005) and Manery (2003) assert that vivid words are particular words that are mainly used in descriptive essay writing. They mainly act as particular modifiers; for instance, the word car in this expression ‘the car’s price is high’ does not seem to be vivid enough. When the word ‘car’ is modified into ‘the red Proton car’, it becomes much clearer because modification adds clarity to the characteristics of the car. Thirdly, the revision and presentation of a fair copy of a descriptive composition in the "post-writing" stage were addressed. The teachers were practically informed how to edit/reedit their peers’ papers, and also gave suitable comments on their peers and their students’ papers not only in the course but also in their own classes.
After the elaboration of self- and peer assessment techniques, the participant teachers were asked to write a composition within the framework of Spencer's (2005) model in the fifth session to assess their understanding after the course administration. The participant teachers were requested again to apply the self-assessment technique in the class. Ultimately, they were assisted to check their compositions.
Finally, the trained Iranian male and female EFL teachers were asked to train and familiarize their learners (i.e., 25 males and 25 females) with Bartlett's and Spencer's models within three months during the summer semester. It is worth noting that to ensure that the process was going on well, we periodically supervised the procedures of used techniques in the classroom.
Instrumentation
To collect the data from both groups of the study, two instruments were developed.
One-to-One-Interview
The first instrument involved seven interview questions concerning self- and peer assessment of descriptive writing, and the questions were formulated based on the administered course. In addition, the interview process helped the researchers elicit the factors and items included in a questionnaire for assessing the EFL elementary learners' attitudes towards descriptive writing course regulated on self- and peer assessment.
Questionnaire
The second tool of the study was a 19-item questionnaire to elicit the elementary EFL learners’ attitudes based on gender differences. The questionnaire’s items were developed based on four finalized themes approved by two EFL experts including, “Cooperation,” “Knowledge,” “Motivation,’’ and “Practice” in the interview stage.
Dependability and Credibility Issues
The dependability and credibility issues of the interviews were assessed by two EFL experts who were Ph.D. holders in Teaching English as a Foreign Language (TEFL) and had more than 15 years of English teaching experience. A pilot study was also conducted based on the newly-designed instrument with a sample of 50 Iranian learners who took part in the course. The participants were able to freely express any ideas about the main structure of the questionnaire and even one-to-one interview form. Various types of feedback (e.g., corrective, evaluative) were also used to modify the instruments so as to enhance the transparency of the items. As an example, some factors were changed, and some items were revised. The findings were then examined by the experts for clarity, relevance, and content consistency. After examining the last version of the questionnaire, Cronbach's alpha test was run on the entire questionnaire and each factor. The obtained Cronbach's alpha values for the factors were as follows: cooperation (.914), knowledge (.892), practice (.942), and motivation (.931). The whole questionnaire also exhibited an acceptable level of reliability, i.e., Cronbach's alpha = .901. Considering Pallant’s (2013) criterion of acceptable alpha value, which is .70, the questionnaire was found to be reliable, and all of the classifications demonstrated acceptable internal consistency and reliability.
Data Analysis
The research involved both qualitative and quantitative phases.
Qualitative Data
In the qualitative phase, primarily, the female and male EFL learners' attitudes toward peer and self-assessment of descriptive writing were taken into consideration, and all were asked to take part in the course. The course aimed to teach the learners descriptive writing within five successive sessions, using Bartlett's (2015) self- and peer assessment model and Spencer's (2005) composition models. The participants were asked to write at least one composition each session. Then, they were asked to participate in a researcher-made one-to-one interview (see Appendix B). After the data reached saturation, the major factors for developing a questionnaire were identified (through the constant comparative method of analysis), and those factors were structured in a five-point Likert scale format. These factors were 19 items that constituted the questionnaire.
Quantitative Data
In the quantitative phase, the same participants were requested to complete the survey form and to express their attitudes toward peer and self-assessment of descriptive writing (see Appendix C). To analyze the data obtained, exploratory factor analysis and one-way MANOVA tests through IBM SPSS Software version 22 were employed.
RESULTS
The results are presented in two different strands. In the qualitative strand, the results of the obtained data for the recommendation and confirmation of the developed the EFL learners' questionnaire’s themes and items are elaborated. In the quantitative strand, the results of the factor analysis and one-way MANOVA are explained, as well.
Qualitative Data Analysis
To address the first research question of the study (i.e., What are the Iranian (female and male) EFL learners' views toward peer and self-assessment of descriptive writing?), the authors recruited the trained EFL teachers to instruct the selected Iranian male and female EFL learners and to elicit their attitudes towards the course. Moreover, to increase the accuracy of the training procedures, all of the EFL learners' compositions in five classes were directly checked. Also, the assigned trainers' suggestions concerning the participants’ compositions were checked and re-checked. Then, with the help of the institutes' principals, the trainers' classes were indirectly supervised and observed via class cameras, and whenever needed, some pieces of advice were given.
Interviews with EFL Learners
All the participating male and female EFL learners were asked to take part in face-to-face interviews so as to express their ideas towards the course. The interviewees’ voices were recorded by means of a voice recorder. It should be mentioned that after administrating 15 successive interviews, data saturation occurred. After that, no new information was elicited from the EFL learners. In addition, the recorded files were transcribed in great detail, and were prepared to be analyzed through the constant comparative method of analysis (CCMA) developed by Strauss and Corbin (1998) within the framework of the grounded theory. As such, four main themes, namely, ‘Cooperation’, ‘Knowledge’, ‘Practice’, and ‘Motivation’ were identified. Finally, to come to an acceptable decision about the accuracy of the identified themes, we analyzed the transcribed files over and over.
Theme Identification in the Learners’ Data
The outcomes of the interview analysis in the qualitative phase of this investigation demonstrated that several factors (including the dimensions in Table 1) were influential in motivating the participants to complete the course successfully. Out of the themes extracted, 25 items were initially selected to elicit male and female EFL elementary learners' attitudes towards the course. These items fell under four main themes: ‘Cooperation’, ‘Knowledge’, ‘Practice’, and ‘Motivation’.
‘Cooperation’ was the degree to which EFL learners were eager to collaborate with one another toward a joint objective or task. For the most part, they agreed that self-and peer assessment were so interesting they were willing to cooperate with one another in descriptive writing. For instance, one of the participants said that “The first and foremost positive point was that our classmates worked cooperatively and enthusiastically within a group and it assisted us to reach a unified idea at the end of the task” (Interviewee no. 1). Participants also expressed many positive aspects of cooperation in assessment. An interviewee stated that “We were able understand each other's errors and also we benefited from our peers' ideas or feedback” (Interviewee no. 9). Another respondent reasoned that “Through cooperation, learners could become sources of learning for their classmates” (Interviewee no. 12).
With regard to ‘Knowledge’, many participants asserted that the information, understanding, and skills that they gained through self- and peer assessment were of great importance. Also, they believed that self- and peer assessment increased their knowledge in descriptive writing. An interviewee reported that “Now I feel that I have enough knowledge to help my classmates in applying self-and peer-assessment in descriptive writing” (Interviewee no. 4). When discussing the assessment knowledge, a few of the interviewees cited knowledge enhancement as an advantage of self- and peer assessment; Interviewee No.15 stated that “Now I feel that I am competent enough to help my classmates in assessing their writing”. Interviewee No. 22 asserted that “I could find my problems and I tried not to repeat them in later compositions”. Interviewee no. 18 also added that “It increased my knowledge of composition writing”.
A related theme that most students referred to was ‘Practice’. It was the application of the techniques that they learned. They believed that through self- and peer assessment, they were able to practice descriptive writing over and over. For example, one interviewee agreed that “Self- and peer assessment give us sufficient chance to practice and learn so many things about descriptive writing” (Interviewee no. 9). Another participant expressed that providing opportunity for practice was another positive aspect of self- and peer assessment. Interviewee No. 20 shared her observation of student active practice in classroom; “It seemed like a great opportunity for students to practice”. Another respondent expressed that “His descriptive writing skill improved through practicing self-and peer assessment” (Interviewee no. 18).
Finally, most of the participating interviewees discussed that ‘Motivation’ was a major characteristic of this type of instruction. It was seen as the driving force that encouraged them to apply the techniques they learned. One participant opined that “What is more, we became acquaintance with each other's comments and problems; and this technique encouraged and motivated us to write a better composition within the framework of peer-assessment only task” (Interviewee no. 7). Respondents also addressed how motivation caused by self- and peer assessment affected interactions with students. Interviewee no. 16 expressed that “motivation created in these types of assessment increased engagement in the class activities related to descriptive writing”. Similar observations were made by other respondents, as well. “Self-and peer assessment had the potential to increase my motivation in descriptive writing”, said Interviewee No. 13.
Quantitative Data Analysis
Likewise, the second research question was addressed (i.e., Do the factors identified predict the Iranian (female and male) EFL learners' views toward peer and self-assessment in the case of descriptive writing?). To answer the second question, the results of the exploratory factor analysis are presented below.
Factor Analysis of EFL Learners’ Data
In addition, the results of exploratory factor analysis (EFA) in the quantitative phase revealed that 4 out of 5 of the confirmed factors were appropriate. Meanwhile, 19 out of the 25 items of the newly-developed scale for assessing the attitudes were acceptable.
Table 1: Descriptive Statistics of the EFL Learners' Questionnaires
|
|
N
|
Min.
|
Max.
|
Mean
|
Std. Deviation
|
Skewness
|
Kurtosis
|
|
|
|
|
|
Statistic
|
Std. Error
|
Statistic
|
Std. Error
|
LI1
|
50
|
4
|
5
|
4.60
|
.495
|
-.421
|
.337
|
-1.900
|
.662
|
LI2
|
50
|
4
|
5
|
4.66
|
.479
|
-.697
|
.337
|
-1.580
|
.662
|
LI3
|
50
|
4
|
5
|
4.62
|
.490
|
-.510
|
.337
|
-1.814
|
.662
|
LI4
|
50
|
3
|
5
|
4.50
|
.544
|
-.396
|
.337
|
-1.052
|
.662
|
LI5
|
50
|
3
|
5
|
4.54
|
.542
|
-.564
|
.337
|
-.885
|
.662
|
LI6
|
50
|
4
|
5
|
4.50
|
.505
|
.000
|
.337
|
-2.085
|
.662
|
LI7
|
50
|
3
|
5
|
4.50
|
.544
|
-.396
|
.337
|
-1.052
|
.662
|
LI8
|
50
|
3
|
5
|
4.50
|
.544
|
-.396
|
.337
|
-1.052
|
.662
|
LI9
|
50
|
2
|
5
|
4.40
|
.639
|
-1.076
|
.337
|
2.434
|
.662
|
LI11
|
50
|
3
|
5
|
4.54
|
.613
|
-.988
|
.337
|
.017
|
.662
|
LI12
|
50
|
4
|
5
|
4.64
|
.485
|
-.602
|
.337
|
-1.708
|
.662
|
LI13
|
50
|
4
|
5
|
4.60
|
.495
|
-.421
|
.337
|
-1.900
|
.662
|
LI14
|
50
|
4
|
5
|
4.52
|
.505
|
-.083
|
.337
|
-2.078
|
.662
|
LI10
|
50
|
4
|
5
|
4.62
|
.490
|
-.510
|
.337
|
-1.814
|
.662
|
LI16
|
50
|
4
|
5
|
4.62
|
.490
|
-.510
|
.337
|
-1.814
|
.662
|
LI15
|
50
|
4
|
5
|
4.66
|
.479
|
-.697
|
.337
|
-1.580
|
.662
|
LI17
|
50
|
4
|
5
|
4.50
|
.505
|
.000
|
.337
|
-2.085
|
.662
|
LI18
|
50
|
4
|
5
|
4.38
|
.490
|
.510
|
.337
|
-1.814
|
.662
|
LI19
|
50
|
3
|
5
|
4.48
|
.544
|
-.315
|
.337
|
-1.106
|
.662
|
LI20
|
50
|
1
|
5
|
2.50
|
.505
|
.000
|
.337
|
-2.085
|
.662
|
LI21
|
50
|
1
|
5
|
2.54
|
.503
|
-.166
|
.337
|
-2.057
|
.662
|
LI22
|
50
|
1
|
5
|
2.62
|
.490
|
-.510
|
.337
|
-1.814
|
.662
|
LI25
|
50
|
1
|
5
|
2.48
|
.505
|
.083
|
.337
|
-2.078
|
.662
|
LI24
|
50
|
1
|
5
|
2.54
|
.503
|
-.166
|
.337
|
-2.057
|
.662
|
LI23
|
50
|
1
|
5
|
2.82
|
1.320
|
.290
|
.337
|
-.937
|
.662
|
Valid N
|
50
|
|
|
|
|
|
|
|
|
Table 1 summarizes the data obtained from all of the participants. The data shows that all of the EFL learners answered the 25 items proposed for designing the EFL learners' questionnaire. Results of skewness and kurtosis statistics also show that the data was distributed normally. Table 1 also shows that the means of the 19 items are greater than 4, while 4.66 is the highest mean, and 2.50 is the lowest mean.
Table 2: Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett's test of Sphericity
|
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.
|
.726
|
Bartlett's Test of Sphericity
|
Approx. Chi-Square
|
12.075
|
Df
|
10
|
Sig.
|
.000
|
Table 2 demonstrates that the obtained Kaiser-Meyer-Olkin Measure of sampling adequacy is .726. It represents the fact that the suitability factor of the questionnaire is acceptable. Bartlett's Test of Sphericity also confirms the adequacy of sampling as the Sig. value is smaller than .05.
Table 3: Communality Measures of the Data
|
Dimensions
|
Initial
|
Extraction
|
CO
|
1.000
|
1.000
|
KN
|
1.000
|
1.000
|
MO
|
1.000
|
1.000
|
PR
|
1.000
|
1.000
|
SWE
|
1.000
|
1.000
|
Table 3 lists the communalities of the data before and after extraction. It shows that the initial communalities of all data are 1. This was due to the fact that the principal component analysis followed the initial assumption that all variances are common. The extraction column in Table 4 clarifies that all of the loaded factors are greater than 0.30, which is the minimum requirement for loading.
Table 4: Explanation of the Overall Variance
|
Com.
|
Initial Eigenvalues
|
Extraction Sums of Squared Loadings
|
Rotation Sums of Squared Loadings
|
Total
|
% of Variance
|
Cumulative %
|
Total
|
% of Variance
|
Cumulative %
|
Total
|
% of Variance
|
Cumulative %
|
1
|
1.487*
|
29.739
|
29.739
|
1.487
|
29.739
|
29.739
|
1.000
|
20.008
|
20.008
|
2
|
1.201*
|
24.023
|
53.762
|
1.201
|
24.023
|
53.762
|
1.000
|
20.006
|
40.014
|
3
|
1.182*
|
19.633
|
73.395
|
1.182
|
19.633
|
73.395
|
1.000
|
20.006
|
60.020
|
4
|
1.115*
|
14.297
|
87.692
|
1.115
|
14.297
|
87.692
|
1.000
|
20.002
|
80.022
|
5
|
.615
|
12.308
|
100.000
|
.615
|
12.308
|
100.000
|
.999
|
19.978
|
100.000
|
N.B. Asterisk represented that components one to four were satisfactory.
Table 4 illustrates the variance of the original variables for every factor. It also shows that only 4 out of 5 components have eigenvalues greater than 1 (1.487, 1.201, 1.182, and 1.115). The components (“Motivation”, “Practice”, “Cooperation”, and “Knowledge”) accounted for 87.692% of the variance.
Table 5: The Total Component Matrix
|
|
Components
|
1
|
2
|
3
|
4
|
5
|
Cooperation
|
.784
|
|
|
|
.576
|
Knowledge
|
.768
|
|
|
.616
|
.509
|
Strengths and Weaknesses
|
|
.441
|
|
.530
|
|
Motivation
|
.465
|
.569
|
.402
|
.536
|
|
|
.763
|
|
.904
|
|
|
Table 5 shows the correlations of the proposed factors. As can be seen, the correlations of the factors, except for factor No. 5, are high. This observation confirms the suitability of the EFL learners' scale.
Table 6: Rotated Component Matrix
|
Practice
|
Components
|
1
|
2
|
3
|
4
|
5
|
Motivation
|
.880
|
|
|
|
|
|
Practice
|
|
.978
|
|
|
|
Cooperation
|
|
|
.899
|
|
|
Knowledge
|
|
|
|
.959
|
|
Strengths and Weaknesses
|
|
|
|
|
.532
|
Based on the rotated component matrix (Table 6), four factors (“Practice”, “Knowledge”, “Cooperation”, and “Motivation”) were considered as the main factors of this newly-developed questionnaire for EFL learners. Yet, given the low correlation of factor No. 5 (“Strengths and Weaknesses”), it was omitted from the final questionnaire, as a result of which items 20-25 were removed. As such, an investigation of the factors revealed that only 19 out of 25 items were suitable for assessing the EFL learners' attitudes.
One-Way MANOVA Administration for EFL Learners’ Data
Moreover, for the third research question, which addresses the differences between the Iranian female and male EFL learners' views toward peer and self-assessment of descriptive writing in terms of the factors identified, a one-way MANOVA was conducted.
The analysis of one-way MANOVA in the quantitative phase demonstrated that the attitudes of Iranian male and female EFL learners were different in terms of gender (see Tables 7-13).
Table 7: Residuals Statistics for Checking Outliers
|
|
Minimum
|
Maximum
|
Mean
|
Std. Deviation
|
N
|
Predicted Value
|
1.07
|
1.80
|
1.50
|
.160
|
50
|
Std. Predicted Value
|
-2.687
|
1.878
|
.000
|
1.000
|
50
|
Standard Error of Predicted Value
|
.075
|
.247
|
.153
|
.042
|
50
|
Adjusted Predicted Value
|
1.09
|
1.78
|
1.50
|
.173
|
50
|
Residual
|
-.718
|
.717
|
.000
|
.479
|
50
|
Std. Residual
|
-1.436
|
1.434
|
.000
|
.958
|
50
|
Stud. Residual
|
-1.492
|
1.595
|
.004
|
1.011
|
50
|
Deleted Residual
|
-.775
|
.886
|
.004
|
.534
|
50
|
Stud. Deleted Residual
|
-1.513
|
1.624
|
.003
|
1.015
|
50
|
Mahal. Distance
|
.129
|
10.980
|
3.920
|
2.640
|
50
|
Cook's Distance
|
.001
|
.120
|
.024
|
.025
|
50
|
Centered Leverage Value
|
.003
|
.224
|
.080
|
.054
|
50
|
|
Table 7 represents data about the presence or absence of univariate / multivariate outliers. Based on the provided data in the table, the maximum value for Mahalanobis distance is 10.98, and it is smaller than the alpha value of 13.82. From the obtained result, it can be concluded that no substantial outliers were found, and this assumption was upheld.
Table 8: Descriptive Statistics
|
|
Gender
|
Mean
|
Std. Deviation
|
N
|
Cooperation
|
Male
|
22.80
|
1.384
|
25
|
Female
|
22.92
|
1.470
|
25
|
Total
|
22.86
|
1.414
|
50
|
Knowledge
|
Male
|
22.12
|
1.787
|
25
|
Female
|
22.92
|
1.152
|
25
|
Total
|
22.52
|
1.542
|
50
|
Motivation
|
Male
|
22.84
|
1.281
|
25
|
Female
|
23.40
|
1.414
|
25
|
Total
|
23.12
|
1.365
|
50
|
Practice
|
Male
|
18.00
|
1.155
|
25
|
Female
|
18.16
|
1.313
|
25
|
Total
|
18.08
|
1.226
|
50
|
Table 8 provides information about the various factors of the dependent variables. For instance, Table 9 shows that the mean values of the attitudes expressed by the males (N=25) on each dimension are 22.80, 22.12, 22.84, and 18.00, while the same values for the females (N=25) are 22.92, 22.92, 23.40, and 18.16.
Table 9: Box's Test of Equality of Covariance Matrices
|
Box's M
|
18.256
|
F
|
1.660
|
df1
|
10
|
df2
|
11015.139
|
Sig.
|
.084
|
As shown in Table 9, the Sig. value is .084, and it is greater than .001. As a result, the obtained covariance matrices of the research variables are similar among the groups. Also, the homogeneity of variance-covariance matrices of the learners’ questionnaires is acceptable.
Table 10: Multivariate Tests
|
Effect
|
Value
|
F
|
Hypothesis df
|
Error df
|
Sig.
|
Partial Eta Squared
|
Intercept
|
Pillai's Trace
|
.998
|
7078.858
|
4.000
|
45.000
|
.000
|
.998
|
Wilks' Lambda
|
.002
|
7078.858
|
4.000
|
45.000
|
.000
|
.998
|
Hotelling's Trace
|
629.232
|
7078.858
|
4.000
|
45.000
|
.000
|
.998
|
Roy's Largest Root
|
629.232
|
7078.858
|
4.000
|
45.000
|
.000
|
.998
|
Gender
|
Pillai's Trace
|
.011
|
1.261
|
4.000
|
45.000
|
.029
|
.011
|
Wilks' Lambda
|
.899
|
1.26
|
4.000
|
45.000
|
.029
|
.011
|
Hotelling's Trace
|
.112
|
1.261
|
4.000
|
45.000
|
.029
|
.011
|
Roy's Largest Root
|
.112
|
1.261
|
4.000
|
45.000
|
.029
|
.011
|
Table 10 shows the data related to the statistical differences between the males' and females' attitudes towards the course. Although the MANOVA test reports several statistics, according to Tabachnick and Fidell (2013), Wilks' Lambda measure should be used for the test result. The observed value (.029) revealed that there is a statistically significant difference between both groups, since the value is greater than .05 – the assigned level of significance in social sciences.
Table 11: Levene's Test
|
|
F
|
df1
|
df2
|
Sig.
|
Cooperation
|
.351
|
1
|
48
|
.000
|
knowledge
|
6.794
|
1
|
48
|
.012
|
motivation
|
.250
|
1
|
48
|
.001
|
Practice
|
.742
|
1
|
48
|
.043
|
Table 11 gives some information about the assumption of equality of error variances across the factors. The obtained data of the table showed that the obtained Sig. levels are .000, .012, .001, .043, and they are smaller than .05. Hence, it can confidently be stated that all of the confirmed factors have equal variances which depicts the fact that the newly-developed questionnaire’s themes are acceptable.
Table 12: Tests of Between-Subjects Effects
|
Source
|
Dependent Variable
|
Type III Sum of Squares
|
Df
|
Mean Square
|
F
|
Sig.
|
Partial Eta Squared
|
Corrected Model
|
COPO
|
.180a
|
1
|
.180
|
.088
|
.008
|
.002
|
KN
|
8.000b
|
1
|
8.000
|
3.540
|
.056
|
.059
|
MO
|
3.920c
|
1
|
3.920
|
2.154
|
.049
|
.043
|
PR
|
.320d
|
1
|
.320
|
.209
|
.009
|
.004
|
Intercept
|
COPO
|
26128.980
|
1
|
26128.980
|
12818.796
|
.000
|
.996
|
KN
|
25357.520
|
1
|
25357.520
|
11220.142
|
.000
|
.996
|
MO
|
26726.720
|
1
|
26726.720
|
14685.011
|
.000
|
.997
|
PR
|
16344.320
|
1
|
16344.320
|
10694.212
|
.000
|
.996
|
Gender
|
COPO
|
.180
|
1
|
.180
|
.088
|
.018
|
.002
|
KN
|
8.000
|
1
|
8.000
|
3.540
|
.016
|
.009
|
MO
|
3.920
|
1
|
3.920
|
2.154
|
.009
|
.043
|
PR
|
.320
|
1
|
.320
|
.209
|
.049
|
.004
|
Error
|
COPO
|
97.840
|
48
|
2.038
|
|
|
|
KN
|
108.480
|
48
|
2.260
|
|
|
|
MO
|
87.360
|
48
|
1.820
|
|
|
|
PR
|
73.360
|
48
|
1.528
|
|
|
|
Total
|
COPO
|
26227.000
|
50
|
|
|
|
|
KN
|
25474.000
|
50
|
|
|
|
|
MO
|
26818.000
|
50
|
|
|
|
|
PR
|
16418.000
|
50
|
|
|
|
|
Corrected Total
|
COPO
|
98.020
|
49
|
|
|
|
|
KN
|
116.480
|
49
|
|
|
|
|
MO
|
91.280
|
49
|
|
|
|
|
PR
|
73.680
|
49
|
|
|
|
|
The tests of between-subjects’ effects indicate that the mean difference across the gender among the four factors is statistically significant. The statistical information in Table 12 reveals a significant difference between males’ and females’ perspectives towards the course.
Table 13: The Marginal Means
|
Dependent Variable
|
Gender
|
Mean
|
Std. Error
|
95% Confidence Interval
|
Lower Bound
|
Upper Bound
|
Cooperation
|
Male
|
22.800
|
.286
|
22.226
|
23.374
|
Female
|
22.920
|
.286
|
22.346
|
23.494
|
Knowledge
|
Male
|
22.120
|
.301
|
21.515
|
22.725
|
Female
|
22.920
|
.301
|
22.315
|
23.525
|
Motivation
|
Male
|
22.840
|
.270
|
22.298
|
23.382
|
Female
|
23.400
|
.270
|
22.858
|
23.942
|
Practice
|
Male
|
18.000
|
.247
|
17.503
|
18.497
|
Female
|
18.160
|
.247
|
17.663
|
18.657
|
As Table 13 displays, the research participants’ attitudes toward the course were statistically significant. As can be easily seen, the females have higher means than the males, which reveals that the female participants expressed more positive attitudes toward peer and self- assessment in the case of the descriptive type of writing.
DISCUSSION
This study sought to explore Iranian EFL elementary learners’ attitudes towards peer and self- assessment of descriptive writing through ESMMsD. To address the study questions, the EFL learners’ interviews were analyzed, and subsequently, exploratory factor analysis and one-way MANOVA methods were employed, sequentially.
The first research question concerned the views of EFL learners toward peer and self- assessment of descriptive writing before and after participating in the course. Based on the obtained data in the qualitative phase, it can be concluded that the participants' attitudes towards self- and peer assessment of descriptive writing were positive for some reasons. First, the results of the interviews revealed that both male female EFL elementary learners had positive attitudes towards the course. The participants were asked to express their views about the course and mention whether they had ever experienced such courses. They admitted their unfamiliarity with the course claiming that they had not experienced it before. After the course was finalized, roughly all participants had positive views towards the course. Furthermore, in the light of the major factors (‘Cooperation’, ‘Knowledge’, ‘Practice’, and ‘Motivation’), as well as the 19 items extracted from the questionnaires, it could be argued that the participants had positive toward the course, and that they were motivated enough to receive training in a similar method in the future.
The second question addressed the factors that could possibly help to predict the Iranian female and male EFL learners' attitudes toward peer and self-assessment of descriptive writing. With regard to the essence of the factors such as (‘Cooperation’, ‘Knowledge’, ‘Practice’, and ‘Motivation’), it could be noted that the participants had positive opinions toward peer and self-assessment of descriptive writing for the following reasons. The first predictive factor is related to the ‘Cooperation’ theme. Besides making the course competitive, the EFL elementary learners’ sense of cooperation was successfully rocketed; that is, their enthusiasms towards accomplishing it simultaneously doubled. The ‘Knowledge’ theme was considered as the second predictive element that the EFL elementary learners attempted to increase towards the course since their familiarization with basic and major principles of self- and peer assessment of descriptive writing ascended. Then, the ‘Practice’ theme was considered as the third predictive factor in operationalizing the participants’ acquired knowledge after taking part in the course it could be claimed that it could have positive impact on the EFL learners’ perspectives towards the course. Moreover, the ‘Motivation’ theme was considered as the last predictive reason that had positive impacts on the EFL elementary learners’ attitudes towards the course because it boosted their motivation.
The last research question of this study addressed the Iranian female and male EFL learners' attitudes towards the course. The results of one-way MANOVA showed that both males and females had positive attitudes towards the course, and their attitudes were statistically significant. More specifically, based on the data in Table 10, the observed Sig. was .029 and represented the fact that males’ and females’ attitudes towards the course were statistically different, as the level of significance was greater than .05. Table 13 further showed that females’ means were 22.920, 22.920, 23.400, and 18.160 sequentially whilst, males’ ones were 22.800, 22.120, 22.840, and 18.000. From the given statistics, it can be inferred that males’ and females’ attitudes toward the course were different from each other and even it can be said that female EFL learners had more positive attitudes towards the course than their counterparts. The results of the current research were generally in line with those of Abolfazli-Khonbi and Sadeghi, (2013), Ashraf and Mahdinezhad (2015), Azarnoosh (2013), Birjandi and Hadidi-Tamjid (2012), Birjandi and Siyyari (2010), Ismail (2010), Maiz-Arévalo (2008), Siow (2015), and Vickerman (2009). Similar to the present study, these assessment researchers have found that learners in general have positive views towards self- and peer assessment.
It should be mentioned that some aspects of this study were considerably different from the previous studies. Most importantly, the present research drew on a new researcher-made self- and peer assessment questionnaire to examine EFL elementary learners’ views. Another major difference could be seen in the combination of the models based on which the course was constructed. This method, which was composed of Bartlett’ and Spencer’s models, was very reliable and could strongly serve self- and peer assessment procedures in descriptive writing. Another key difference of this study was its use of the sequential exploratory mixed-methods design, as previous research focused only on qualitative or quantitative designs. The present application could highlight the novelty of the research and the systematic results and findings it provided. Finally, although previous research treated writing skill as a whole, this study focused on one particular genre of writing. This choice facilitated the administration of the course, while keeping the participants’ attitudes completely focused on one well-defined genre. In short, the study showed that EFL learners’ increased exposure to self- and peer assessment in class could foster more positive attitudes towards this mode of assessment in the case of writing skill.
CONCLUSION AND IMPLICATIONS
demonstrated that the attitudes of the female and male elementary EFL learners toward peer and self-assessment techniques in descriptive writing were significantly different. While taking part in the course, the participants attempted to include their personal experiences and information learnt from assessing their own or their peers' compositions.
From both theoretical and practical perspectives, the study offers some contributions which are as follows. To be more specific, the study contributed to the theoretical knowledge of self- and peer assessment of descriptive writing in an EFL context. Based on constructive principles, self- and peer assessment support active and cooperative learning which could lead to learner motivation and increased learning outcomes. A social and educational benefit related to self- and peer assessment in this study was learner engagement in cooperative tasks during feedback exchanges, pair work, and discussion.
Moreover, the outcomes of this study have practical implications for those involved in language assessment. Considering the positive attitudes of the study’s participants, this investigation could also enhance assessment researchers and practitioners’ understanding into self- and peer assessment. In fact, the findings based on EFL learners’ perspectives suggested that Bartlett’s (2015) and Spencer’s (2005) models offered practical methods for assessing descriptive writing.
This study was conducted based on traditional self- and peer assessment. With the rise of online education, future studies need to focus on more recent types of self- and peer assessment, such as online peer assessment, or rather e-peer response. To shed more light on the feasibility of peer and self-assessment in other EFL settings, a few limitations that we observed during the study, should receive serious attention; for instance, (a) cultural aspects might discourage learners to make comments and provide feedback, or (b) some students may have a preference for teacher feedback. Future research could doubtlessly continue to explore in more detail how these issues can be dealt with.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID